Nonlinear association between lung needle path CT attenuation values and postprocedural immediate pneumothorax following computed tomography-guided lung biopsy: a retrospective cohort study

Table 2 Linear and two piecewise linear regression model

Models	Immediate pneumothorax ratio (OR, 95% CI, p-value)
Model 1
Model fitted by linear regression	0.98 (0.98, 0.99) < 0.001
Model fitted by two-piecewise linear regression on the infection point of the lung needle path CT attenuation, (Hu)
≤ -805	0.98 (0.97, 0.99) < 0.001
> -805	1.00 (0.99, 1.01) 0.879
p-value for the log-likelihood ratio test	0.003
Model 2
Model fitted by linear regression	0.99 (0.98, 0.99) < 0.001
Model fitted by two-piecewise linear regression on the infection point of the lung needle path CT attenuation, (Hu)
≤ -805	0.98 (0.97, 0.99) < 0.001
> -805	1.00 (0.99, 1.01) 0.748
p-value for the log-likelihood ratio test	0.001
Model 3
Model fitted by linear regression	0.99 (0.99, 1.00) 0.013
Model fitted by two-piecewise linear regression on the infection point of the lung needle path CT attenuation, (Hu)
≤ -805	0.99 (0.98, 0.99) 0.001
> -805	1.00 (0.99, 1.01) 0.566
p-value for the log-likelihood ratio test	0.028

Model 1: no adjustment; Model 2: adjusted for age, gender, height, weight, chronic obstructive pulmonary disease, alcohol consumption, smoking status, systolic blood pressure, and diastolic blood pressure; Model 3: adjusted for model 2, plus lesions located in the lungs and lobes, lesion diameter, tract length in the lungs, needle-pleura angle, puncture through the interlobar fissure, number of needle redirections, number of pleural punctures, puncture needle gauges, operation duration, pulmonary hemorrhage along the needle path, operator experience, and patient puncture position. CI, confidence Interval; Hu, Hounsfield units; OR, odds ratio

ISSN: 1471-2466